The Role of Surveillance Imaging for the Detection of Relapsed Lymphoma.

Up to 30% of patients with Hodgkin's lymphoma (HL) and 60% of patients with aggressive Non-Hodgkin's lymphoma (A-NHL) will relapse after first remission. Early detection of relapse, associated with a low tumor burden, may improve survival. The optimal follow up method - clinical versus imaging - for the detection of relapse has not been clarified.

We retrospectively reviewed the files of 125 patients with HL and A-NHL diagnosed between 1.1993 - 1.2009 who relapsed at least one month after the end of initial therapy. We assessed whether relapse was detected by clinical symptoms or by imaging and specifically queried whether recent imaging techniques i.e. [¹⁸F] fluorodeoxyglucose positron emission tomography (FDG-PET), have changed the pattern and outcomes of relapsed disease.

Forty two (34%) patients had HL and 83 (66%) had A-NHL. Of the 125 patients, age was <30 years in 50% of HL and >60 years in 47% of A-NHL patients; 75 (60%) had advanced disease and 50 (40%) had early disease at diagnosis. Seventy patients (56%) relapsed in the first year following treatment, 20 (16%) in the second year and the rest (33, 26%) relapsed thereafter. In 75 (60%) patients, relapse was detected based on patient's symptoms or an abnormal physical finding (clinical detection of relapse, CDR) and in 50 (40%) patients relapse was detected by routine imaging (image detection of relapse, IDR). A significantly higher proportion of A-NHL patients had CDR as opposed to IDR (67% versus 33%), whereas in HL the opposite was found (45% CDR and 55% IDR) (p=0.022). In the years 2001-2009, when FDG-PET was available at our institution, 28% of HL patients had CDR and 72% were diagnosed by IDR (p=0.065). Multiple regression analysis confirmed the independent effect of histology (OR 2.6, 95% CI 1.19-5.69 for HL versus NHL) and period of relapse (OR 2.4, 95% CI 1.1-5.4 for ≥2001 versus '2000) on the probability of IDR. Characteristics at relapse including time to relapse, stage, presence of B symptoms, prognostic score, site of relapse, extranodal involvement and period of relapse did not influence the mode of diagnosing relapse. The overall survival after relapse of all 125 patients did not differ significantly whether they had CDR or IDR. However, in patients with HL (but not in patients with A-NHL), IDR was associated with improved survival (albeit non-significant) (Fig 1). The risk of death was more than twice for HL patients with CDR versus IDR (Hazard ratio (HR) 2.4, 95% CI 0.73-7.63) whereas in A-NHL patients the mode of detection of relapse was not associated with survival (HR 0.91, 95% CI 0.5-1.66).

These preliminary results show that the major mode of detecting relapses in lymphoma remains the clinical exam and not imaging. However, routine surveillance by imaging may be important in HL as opposed to A-NHL and the use of contemporary imaging modalities, i.e. FDG-PET, may be more sensitive for relapse detection. Although non significant, our findings suggest that image detection of relapse in patients with HL may be associated with improved survival.

View largeDownload slide

View largeDownload slide

Close modal

No relevant conflicts of interest to declare.